Nowadays, data are easily obtained from everywhere. However, the problem of confidentiality or privacy of information of the data becomes important because the information can be extracted from the data such as using data mining, which sometimes may inadvertently divulge such information. In order to protect the privacy of data and also guarantees precise information extracted in accordance with the original information, thus using privacy-preserving data mining (PPDM). This study proposes a hybrid transformation in PPDM, which is a merger of the two existing techniques on previous studies, the entropy-based partition technique and combined distortion techniques. To measure the proposed method, evaluation of the utility and privacy parameter evaluation are used. Utility evaluation is used to assess the accuracy of the information and privacy parameter evaluation to assess how close the original value will be obtained from the transformation and how much they are distorted. The experimental results show that the proposed method gives better results than previous methods in utility and privacy, so the data will be preserved and can be used for analyzing such as data mining.
展开▼
